RNG testing Thrall perks

[DanQuixote]

tl:dr RNG seems to be working pretty much as it should be - pretty much the only thing I think I’ve really confirmed is the existence of confirmation bias…

I’ve done some more testing - this time how the thrall perks rng seems to be turning out. Long story short, I levelled several Cimmerian Berserkers in game and wasn’t impressed by their perks so wanted to see how it would turn out over a larger sample size.

Note: perks (and even stats) are relatively unimportant on thralls. Many of us know this. It doesn’t stop it being frustrating/disappointing to spend hours leveling a thrall only to see a negative perk come up, and it is especially frustrating for newer players who don’t always realise how little it really means overall. There is a common perception (myself included) that better thralls seem to get worse perks. This test doesn’t disprove that theory, but the results do suggest against it, at least to me.

Methodology -
I spawned in 100 Cimmerian Berserkers, placed them, admin levelled them to 20 (without food) and then collated the results. This is not a large enough sample size (but it would need tens of thousands to really smooth out the results) - but the results do suggest to my eye that overall perk distribution is pretty fair, and actually better than we tend to think.

This first table shows the complete tally of all perks received by the 100 berserkers, divided in columns of 10 thralls at a time.

The second table is more subjective, but does reflect the way many players tend to think about perks (though perks like archer or eagle-eyed can also be felt as negative, but would not count as such under this tally system as they do not produce an actual minus to strength or vitality):

Finally, the most subjective of all, I tried to view the sets of perks from the perspective of how I would feel about that set of perks in game. The result surprised me, with 47 thralls with ‘good’ perks, 21 with ‘bad’ perks and 32 ‘eh, whatever’ perks.

For examples of how I categorised this - a thrall with High Strung, Near-Sighted and Angry Drunk counted as ‘good’ perks (despite only having one ‘positive’ by the previous test), whereas I rated a thrall with Unflinching, Archer and Eagle-Eyed as ‘bad’, because that set feels disappointing on a fighter (despite there being no negatives by the previous test).

A few other interesting results -
Assassin, Strong Arms, Redeemed
Vanguard, Redeemed, Godly Vigor (80 vitality total)
Vanguard, Cannibal, Angry Drunk
Fit, Redeemed, Might of Giants

And, as can be seen from table 2, 17% fit somewhere around this - though, obviously, not all positive perks are created equal.

What this test doesn’t (and can’t) take account of is the way a single ‘bad’ perk feels - when you’ve spent a few hours to get them to their next perk and it turns out negative, that can colour a players perception of the whole set of perks. But looked at over a large enough sample size, it doesn’t really look as bad as all that.

For anyone wanting to access a PDF version of the tables

For anyone wanting to access a word doc version of the tables

[TeleTesselator]

Now, there’s a bit of science you don’t see everyday…

Good work bro!

[Taemien]

@DanQuixote

Note: perks (and even stats) are relatively unimportant on thralls. Many of us know this. It doesn’t stop it being frustrating/disappointing to spend hours leveling a thrall only to see a negative perk come up, and it is especially frustrating for newer players who don’t always realise how little it really means overall. There is a common perception (myself included) that better thralls seem to get worse perks. This test doesn’t disprove that theory, but the results do suggest against it, at least to me.

Its for this reason we should be focusing on hammering Funcom to change the situation on hidden stats. They need to be removed. What you see should be what you are getting.

Can you imagine a newbie being told to take a named thrall to deal with a situation, take a T4 Exile with the godly of godly perks and watch it get smushed in three attacks? When an unnamed Relic Hunter Treasure Seeker with crap perks would trounce through anything they throw it at.

I admire the work you have done with these tables. I really do. But I fear any sort of focus on perks may cause a stance of “Oh we’ll just adjust Fighters to only get non-Archer perks and vice-versa” and call it a day. That would suck.

We have to drive out the hidden stats first and foremost.

[DanQuixote]

@Taemien

We have to drive out the hidden stats first and foremost.

I don’t disagree that they need to be outed - I seem to wind up needing to explain the hidden modifiers quite often. But I’m not sure how much re-work is actually needed - simply unhiding the modifiers would pretty much solve the actual problem part (so long as it was displayed sufficiently prominently). So long as the players have direct in-game access to the information, I’m not so worried by the system itself.

@Taemien

take a T4 Exile with the godly of godly perks and watch it get smushed in three attacks

As it goes, I often take t4 exiles into inappropriate situations (probably got the habit back when even exiles were pretty godly), and recently took Tarman boss hunting on barbaric - he did fine, despite the fact it turned out he was still in non-epic armor (I’d forgotten to upgrade). However, your point is entirely legitimate, and I’ve seen players quit after losing one too many ‘good’ thralls (that weren’t really…).

So, I agree with the importance of focus on the hidden modifiers, but I also feel that the disappointment over ‘bad’ perks can have considerable impact and is worth also being a part of the focus.

(And while we’re on the subject of re-working followers - a ‘attack no one - flee’ behaviour (like the horse already does) would be useful as well - but simply displaying the ‘hidden modifiers’ would be a (hopefully) simple change that could make a big difference for a start.)

[Taemien]

The reason I want to see them gone is because they affect more than damage.

You have the modifiers for Ranged and Melee. But there’s also modifiers for Vitality to Health and Agility to Armor. I believe Survival buff reduction is the same across the board.

But I would like to see the modifiers removed. I would then like to see them get the SAME damage increases from STR/ACC as we do. The same Armor from AGI as we do. Vitality is an interesting one since thralls do need a bit more health to survive than players. But this one should be standard across the board. So a 20 Vitality thrall will have less health than one with 40.

This way a high Strength thrall you know will hit harder. A higher Vitality thrall will have more health. One with more Agility will have a bit more armor. Then the Tier of the thrall determines how high those stats start at.

So the idea would be a 0 Strength Dalansia with an Iron Weapon does the same damage as a 0 Strength Tier I exile with the same weapon. Of course she would never come with 0 strength because of being from the Forgotten Tribe and being a named.

THAT would make it better for new players and old. And the perks will be useful.

Hell you could end up in situations where a lower tribe’d named thrall could be better if they get the good perks. That would feel nice. Course bad perks could make things bad, but then we can have that discussion.